Pharmacogenetics research has resulted in deeper understanding of the etiology of variability in drug metabolism and response, including antiretroviral drugs in HIV. Although genetic influences on drug clearance rates (pharmacokinetics) have been established clearly in a few circumstances, the relationship between genes and treatment response (pharmacodynamics) has been less clear-cut. The objective of this proposal is to determine if polymorphisms associated with slower metabolism of antiretroviral drugs result in better or worse treatment outcomes. Given variable adherence amongst individuals in clinical settings, slower drug metabolism of individual drugs may make those drugs more forgiving of lower adherence for those individuals and therefore result in greater treatment success. Alternatively, the higher concentrations in slow metabolizers may result in intolerable toxicity, non-adherence, and thus treatment failure. Or, the prolonged half-life in slow metabolizers who temporarily stop all of their medications may result in exposure to the slowly metabolized drug alone- functional monotherapy- which may cause emergence of resistance, and treatment failure. We aim to determine whether subjects with slow drug metabolizing genotypes have 1) higher or lower rates of virological suppression and 2) lower or higher rates of virological breakthrough than those with other genotypes. In addition, we aim to determine if the relation between genotype and the 2 treatment outcomes is modified by the individual's medication adherence. We will conduct two studies in HIV infected individuals on efavirenz-containing regimens in one of the largest HIV treatment sites in the world in Gaborone, Botswana. The primary exposure will be the genotype of the cytochrome P450 2B6 enzyme, which metabolizes efavirenz. The first study will be a prospective cohort of treatment naove individuals initiating efavirenz-containing regimens with the outcome being viral load at 6 months. The second study will be a case-control study with cases being individuals who had achieved undetectable viral loads on efavirenz at 6 months, but subsequently had virological failure. Cases will be individuals in the same clinical cohort who do not have virological breakthrough after achieving undetectable viral loads. We will assess the effect of adherence by testing for a genotype x adherence statistical interaction. If the hypotheses are confirmed, future work will assess other drug-genotype relations and test whether strategies utilizing human genetic testing for tailoring treatment choices would result in better HIV treatment outcomes. If outcomes are improved, given the life-saving nature of antiretroviral therapy, this research has the potential both to prolong survival of individuals and prolong the efficacy of medications for the HIV infected population of the world.